im an alien of exceptional ability

Already some opinionated ppl are anti-comments in code since they inevitably go out of date and the code is of course ground truth. Perhaps comments should now be an optional UI overlay in your code editor, generated by an LLM but not actually part of the source code

Let’s go ahead and reserve the word “decompress” for when LLM researchers take a vacation. Taking a lil break by driving perplexity through the roof

Our team OpenAI is hiring technical staff to work on *midtraining* research in London. Yes, pretraining and post training had a baby. Get in touch if you want to collaborate closely together on efficient LLM architecture. I just moved back! 🇬🇧 openai.com/careers/resear…

A common pattern for me is to like a tweet then mute the account. “You have engaged me, thank you. I seek no further engagement at this time”. Probably why my “for you” feed is full of totally random shit

OpenAI huge shout-out to the fine-tuning eng team jhallard Andrew Peng will hang and Scott Cory for sprinting all weekend to get this launched just days after the gpt-4o-mini launch

We’re testing SearchGPT, a temporary prototype of new AI search features that give you fast and timely answers with clear and relevant sources. We’re launching with a small group of users for feedback and plan to integrate the experience into ChatGPT. openai.com/index/searchgp…

It's only a fair number if I'm evaluated under chain-of-cigs conditions

As someone who has spent way too much time thinking about citations, I am super excited about what Sid and team are building for financial analysis w/ Finster AI. Love to see LLMs synthesizing tons of information and giving provenance for claims 🔥 (Disclosure: I angelled)

Where is there the most alpha in AI research right now?

Which deep generative model kicked off the generative AI era?

Analysts often track metrics like Net revenue retention (NRR) for SaaS companies that aren't listed in an income statement or balance sheet and require manually going through filings and earnings calls. With Finster, you can use a single instruction to extract these metrics

Come check out o1-mini: SoTA math reasoning in a small package openai.com/index/openai-o… with Hongyu Ren Shengjia Zhao Eric Wallace & the rest of the OpenAI team

I *didnt’t* work on our groundbreaking new reasoning model. But OpenAI is the kind of company where I feel a part of it anyway, and feel like my teammates’ success is my own. Come join us if you want to work in an environment like that!

We’re hiring in SF and London! Come join this dumpster fire :)